
This study investigates rewriting practices and learner perceptions in a digital storytelling project involving Chinese university students of French as a foreign language and the use of artificial intelligence. Using an exploratory mixed-methods approach, it draws on questionnaire data, semi-structured interviews, and a qualitative analysis of five collaboratively written scripts. It examines how learners moved back and forth between scriptwriting, filming, post-filming revision, and voice-over recording, as well as how they used AI during linguistic revision. The participants were thirteen B1–B2 students at a university in southern China. They viewed the iterative writing process positively and reported gains particularly in vocabulary, together with greater attention to narrative organization. Ratings were somewhat lower for grammar and textual organization. The analysis of the scripts identified linguistic and discourse choices that reflected the relationships among text, image, voice, and the communicative purposes of the productions. The findings point to the need to embed explicit language-focused instruction at points when linguistic needs arise during the project and to help learners critically evaluate, verify, and adapt AI-generated suggestions.
This article presents an exploratory study comparing a guided oral practice intervention integrating ChatGPT Voice with a control intervention conducted without the tool. The study focused on the pronunciation and grammatical values of the word plus among twelve adult learners of French as a foreign language at B1 level. Both groups followed the same instructional progression, focusing on the phonetic realizations [ply], [plys], and [plyz] in negative, quantitative, comparative, superlative, and liaison contexts. Three assessment tasks – reading aloud, an oral question-and-answer task, and a written test – examined the phonetic realization of plus and its use in context. Results show progress in both groups in the reading-aloud task, suggesting the role of the shared instructional progression. The experimental group, however, displays a more consistent profile in the interactive oral task, particularly in contexts where final /s/ is required. Written performance remains similar across the two groups. Self-reported data indicate that the tool was perceived as useful for structuring oral production, encouraging repetition, supporting reformulation, and drawing attention to the relationship between phonetic form, grammatical context, and meaning. The study suggests that the tool may support guided oral practice, provided that it is embedded in an explicit pedagogical framework.
This article analyses the Massive Open Online Course (MOOC) Didactique du FLE dans une perspective francophone, developed by the University of Geneva and the International Organisation of La Francophonie (OIF) for teachers of French as a foreign language (FLE). It examines the extent to which the programme helps to raise the profile of educational practices that are often marginalised and to strengthen the sense of belonging to a transnational professional community. The analysis employs a mixed-methods approach, combining several sources of data – statistics on progress, badges and course completion rates, post-MOOC questionnaires, forum discussions, tutor feedback and participant testimonials. The results indicate that the contextualised pedagogical approach, the open and inclusive Moodle platform, human tutoring and the progressive recognition of learning are associated with sustained engagement and perseverance, as well as the development of a transnational learning community. Thematic forums and human tutoring appear, in particular, to help highlight diverse professional experiences and to support pedagogical reflection. This MOOC/CLOM is thus an example of an experiment with an inclusive, contextualised and modular programme, the effects of which merit further confirmation through additional research.
[Résumé en anglais] This study offers a synchronic overview of the uses of generative artificial intelligence (genAI) in secondary-level language education. Drawing on an empirical analysis of 50 teaching scenarios published on Édubase between 2022 and 2025, it investigates how genAI is integrated into pedagogical practices through the theoretical frameworks of co-agency and AI literacy. The analysis identifies three main configurations of use: the teacher alone using genAI as a design and assessment assistant; the learner alone using it as a tutor or language partner; and a more collaborative teacher–learner–genAI configuration, in which AI also becomes an object of shared reflection. Even if teachers remain the main drivers of pedagogical intentionality, co-agency frequently emerges through forms of critical regulation exercised by teachers and/or students, particularly via the evaluation of genAI-generated outputs and the iterative refinement of prompts. Findings indicate that approaches to AI literacy predominantly emphasize critical and technical skills – especially prompting – while paying comparatively limited attention to understanding the underlying functioning of AI models and to fostering reflexivity regarding reading, writing, and learning processes. The article thus highlights levers for positioning genAI as an empowering pedagogical partner, based notably on the preservation of human responsibility and the reinforcement of disciplinary knowledge.